Oral Health-Head and Neck Cancers: Addressing Confounding Through Negative Control and Quantitative Bias Analyses.

Area of Science:

• Epidemiology
• Oncology
• Oral Health Research

Background:

• The association between oral health and head and neck cancers (HNC) is biologically plausible but yields conflicting study results.
• Concerns exist regarding mediators, unmeasured risk factors, and biases influencing observed associations.
• Negative control exposure and quantitative bias analysis (QBA) were employed to address confounding and misclassification.

Purpose of the Study:

• To evaluate if unmeasured confounding influences the association between oral health and HNC risk.
• To estimate the extent of bias from non-differential exposure misclassification using QBA.

Main Methods:

• A hospital-based case-control study (HeNCe) in Montreal, Canada, involving 389 HNC cases and 429 controls.
• In-person interviews collected life course exposure data; unconditional logistic regression analyzed associations.
• Sexually transmitted diseases (STD) served as a negative control exposure; QBA assessed misclassification bias.

Main Results:

• Complete denture use (OR=1.33) and >9 missing teeth (OR=1.31) showed increased HNC risk.
• Negative control analysis revealed no significant bias from unmeasured confounders.
• Bias-corrected estimates indicated a stronger association between oral health indicators and HNC risk.

Conclusions:

• Unmeasured confounding did not impact the oral health-HNC risk association.
• QBA-corrected estimates suggest crude associations may underestimate the true magnitude of risk.